石琳 任超凡 于涛李江鹏
(1.内蒙古科技大学数理与生物工程学院内蒙古包头014010;
2.内蒙古科技大学信息工程学院 内蒙古包头014010）
摘要：在高炉炼铁过程中，铁水硅含量是表征炉温热状态的主要参数指标。本文利用包钢6#高炉2011年连续2个月的铁水硅含量700炉生产数据，将金融领域中预测股票波动的时间序列模型用于高炉铁水硅含量的预测中，建立了铁水硅含量的时间序列预测模型。该预测模型重点考虑了炉温的波动性﹑非对称性﹑异方差性，克服了以往炉温控制模型只针对炉况较稳定下才能预测的缺陷。因此该炉温模型预测命中率达到80%，取得较好的预测效果。
关键词：TGARCH模型；炉温波动性；非对称性；异方差性；时间序列
Prediction of hot metal silicon content TGARCH model research
SHI Lin1, REN Chao-fan2,YU Tao1,LI Jiang-peng1
(1. Mathematics, Physics and Biological Engineering School , Inner Mongolia University of Science and Technology ,Baotou 014010, China; 2.InformationEngineeringSchool;Inner MongoliaUniversityof Science and Technology,Baotou014010,China;)
Abstract:The hot metal silicon content is the main parameter for the characterization of the thermal state of furnace temperature in the iron making process of blast furnace.The time series model of forecasting the stock fluctuation in the financial field was adopted for the prediction of blast furnace hot metal silicon content , and the time series model of forecasting the silicon content in hot metal was established using the content of hot metal silicon production data by 6 # blast furnace of Baotou Steel in 2011 for two consecutive months after 700 times’ production in the furnace. The prediction model strongly emphasized on the furnace temperature fluctuations, non-symmetry and heteroscedasticity, overcome the defect of the temperature control model of the previous furnace that could only forecast under more stable furnace condition.Therefore, thehit rateof thispredictionmodelreached80%,with better forecast resultsachieved.
Keywords:TGARCH model. Furnace temperature fluctuation; Non-symmetry;Heteroscedasticity;Time series